Why is Data Science Necessary? Why Now?

Why is Data Science Necessary?

Jet Brains Python Developer Survey (2021)
- 52% use python for Data Analysis
- 37% use python for Machine Learning
29% consider themselves to be Data Scientists

Why is Data Science Necessary?

What is Data Science?

Defining things is hard
Two words
1. Data
1. Science

What is Science?

Science is the study of living things and non-living things.

Meaningless

: Where is Hassan?

: He is in the Universe.

What does science mean to the general public?

Scientific equipment
Setting (Laboratory)
Equations?
Scientist

Historically we have known about the world through our 5 senses.

: I know what you are thinking.

And interpreted by our brains.

%%{init: {'themeVariables': { 'fontSize': '1.5em'}}}%%
flowchart LR
    Senses --> Brain
    Brain --> Insight

Our senses

Sight
Smell
- Gas Molecules
Taste
- Liquid Molecules
- Chemical sensing, just like taste

Hearing
- Pulsations of the air
Touch
- Pressure & Temperature sensing by the skin

A loop hole in the sense of touch

Temperature

Goal: Minimise the temperature difference. Touch something at 36°C/37°C (body temperature)
Pressure

Goal: Minimise pressure at point of contact

What are we going to touch?

Solid
- hard to regulate temperature
- you have to be very gentle
- no suprise factor
Gas
- Hard to contain
Liquid
- Easy to contain
- Easy to regulate the temperature

Steps

Get a cup full of water to 37°C
Close your eyes and gently lower your fingers straight down towards the surface of the water.

Equipment & Setup

Observation

When your fingers brake the water surface, you will not feel it.

Senses are not enough

We cannot even calibrate them

%%{init: {'themeVariables': { 'fontSize': '1.5em'}}}%%
flowchart LR
    Brain --> Insight

Greeks

2600 years ago
- Had enough time to think
- When a world starting from what they could observe
- Start of Philosophy
Disagreements
After 200 years
- Logic (Valid Thinking)
- No Disagreements

From the Greeks

To the Romans
To the Early Islamic Civilisation
- Key influence from Indian sources

Western Europe

15-17th centuries, rehabilitated Philosophy by borrowing from logic and mathematics
Ideas are only means to conclusions you can prove
Natural Philosophers (Scientists)
Scientific Method

Then, in the 14th century something uninteresting happened. Western Europe found itself wth had stable societies with a large enough class of people who did not have to work with their hands and they inherited this whole collection Logic, Mathematics and Philosophy (with its many conflicts). This time it was different. Over the 15-17th centuries, Western Europe started rehabilitating philosophy. Borrowing methods from Logic and Mathematics that the only permissable conclusions are those that have been proved.

Ideas are only a means to conclusions that you can prove. The practioners of this revival in Philosophy called themselves Natural Philosophers. Today we know them as scientists. This system of careful study and communication of results is what we call the Scientific Method.

This transformational power of the scientific method is responsible for the world we live in. Science is not Biology, science is not Physics, science is not Chemistry, science is not mathematics, those bodies of work are just by-products of the discipline.

So we have an answer to our question.

What is Science?

Science is the practice of evaluating evidence using sound reasoning to gain insight.

Scientifically Literate

What is Data?

Recorded information

What is Data Science?

The practice of evaluating recorded information with sound reasoning to gain insight

Why Now?

Using Google Ngrams

But Data has been around for along time, how come the practice of data science is new. We turn to tool to help us out.

That tool is Google Ngrams. Google scans books into its digital library and it does some statistics on the words in those books. For example, the number of times a word or a sequence of words appears in a year. Those are what ngrams are. We are going to track the term “Data Science”.

We see here that 1960 has no “Data Science” and towards 2020, there is an explosion which started around 2011.

Let us ignore that explosion and look at 1960 - 2011. Overall, we see Data Science trying to take off and failing. And from the literature we know why.

Data Science starts becoming a thing when computers become products that institutions could buy in the early 1960s. Towards 1970, the computers fail to live up to the hype. Researchers take a step back. There is renewed hope in 1980 but it only lasts until about 1984 and disappointment once again. These periods of disappointment are term as “AI Winters”.

In the early 1990s the World Wide Web is invented and promises a new beginning, but in 2000 web companies collapse. Then it picks up around 2003, but there is affected by the financial crisis of 2007.

In 2011 computers became powerfull enough to run the interesting Data and AI algorithms. And as a result Data Science seemed limitless. So, the explosion.

Why Now (for us)?

Is it because we have a lot of data?

Thank You